Lead response
Immediate service on WhatsApp or the website, with handover to the team.
AI marketing AI applied to the marketing your company already does.
Not a course or a standalone tool. It's implementation: AI in campaign analysis, lead qualification and content, with a success metric agreed before starting.
The service
What AI marketing is, in practice.
For a company that already advertises and serves customers, the useful question is: in which part of the process does AI save time or improve a decision, and how do you know it worked?
AI in paid media
Reading search terms, comparing creatives, text variations for testing and a summary of what changed during the week. Budget decisions stay human. AI shortens the path to them. This is the part closest to paid traffic management.
AI in lead qualification
The ad lead is answered right away, with qualifying questions, and reaches the team organized. The most common case is the AI WhatsApp chatbot.
AI in content
Drafts of posts, scripts and FAQ answers, always reviewed by someone who knows the business. Unreviewed generated content tends to sound generic and may contain wrong information.
Marketing automation with AI
Follow-up of leads who didn't reply, CRM organization and alerts for the team. Small automations, one at a time, each with an owner.
Use cases
Where AI usually comes in first.
Start small, measure and only then expand.
Campaign reports
A summary of what was spent, what came back and what changed, ready for the decision conversation.
Ad variations
More text versions to test, with human approval before going live.
Content and SEO
Research into customers' real questions and drafts that become reviewed pages and articles.
CRM organization
Lead classification, conversation summaries and follow-up reminders.
Internal routines
Repetitive marketing team tasks that eat up hours and don't require judgment.
Comparison
Course, tool or implementation?
Each solves something different.
| Criterion | AI course | Subscription tool | Implementation |
|---|---|---|---|
| What it delivers | Knowledge for the team | Access to a resource | A working process, with a metric |
| Who does the work | The team, after the course | The team, on its own | Specialist and team, together with the company |
| Integration with ads and CRM | No | Depends on the tool | Designed in the project |
| Common risk | Nobody has time to apply it | Unused subscription | Scope too large (that's why it starts small) |
The process
How AI implementation in marketing works.
From diagnosis to follow-up, with one metric per step.
Diagnosis
Where time is lost, where leads go cold and which tools the company already uses.
Priority
Choosing one use case, with the metric that will tell whether it worked.
Pilot
Small-scale implementation, with human review of what the AI produces.
Adjust and expand
What worked scales up. What didn't is dropped without drama.
Follow-up
Periodic review of the numbers and of answers and automations.
Who leads
AI connected to someone who understands media and sales.
Few people combine both: paid traffic and AI implementation. Ramon Saium leads the strategy and connects AI to what already happens in campaigns and customer service. A qualified team handles implementation under his direction.
If you've already paid for an AI solution that sat unused, the conversation starts there. Scope, metric and exclusions come in writing.
Investment
How much AI marketing and AI consulting cost.
There's no standard price list. Waxi itself says robust public research on AI consulting prices in Brazil is lacking. The ranges below show the most common models.
Brazilian market reference, Oct/2026 Ranges in Brazilian reais (BRL) published by other companies in Brazil. These are not ArcWords prices.
| Type of engagement | Market range (BRL) | What it usually includes | Source |
|---|---|---|---|
| Initial diagnosis | R$ 1,500 – 5,000 | Survey of processes and priorities. | Menthoria |
| Pilot or one-off project | R$ 3,000 – 15,000 | One use case implemented, usually in about 30 days. | Menthoria |
| Monthly follow-up | R$ 2,000 – 8,000/mo | Assisted operation and ongoing adjustments. | Menthoria |
| AI automation, CRM and funnels (agency) | R$ 3,000 – 8,000/mo | Range published by an agency in São Paulo. | Agência 365 (SP) |
| Corporate AI consulting | R$ 10,000 – 40,000 | Diagnosis and prioritization in 3 to 6 weeks, author's estimate. | Waxi |
| Corporate assisted operation | R$ 5,000 – 20,000/mo | A different tier: complex data and integrations. | Waxi |
Licenses, APIs and cloud are outside these ranges, according to Waxi. Ranges reviewed quarterly.
What changes the price
Number of use cases
A focused pilot costs less than several processes at once.
Integrations
Connecting CRM, ad platform and WhatsApp increases the technical work.
Data quality
Messy data needs cleaning before any AI works well.
Usage volume
More conversations, content or reports mean more API usage.
Team training
If the team will operate alone afterwards, the project includes training.
Billing model
Diagnosis, fixed project, monthly fee or a hybrid format.
Usually included
- Diagnosis and choice of use case
- Flow and success metric design
- Implementation and tests
- Human review of what the AI produces
- Guidance for the team to operate it
Not included, paid separately
- AI tool subscriptions
- AI API usage (tokens) and cloud
- CRM and automation licenses
- Ad media budget
ArcWords sends a proposal after the diagnosis, with scope, deliverables and exclusions in writing.
Metric
How to know whether AI in marketing is working.
Before starting, agree on the number that will say whether the project worked. Without it, any result becomes opinion.
- Customer service: time to first response and how many leads reach the team qualified.
- Media: hours spent on reports and how quickly what doesn't work gets cut.
- Content: volume published with approved quality, without growing the team.
- Sales: cost per qualified lead, which combines media and customer service.
What AI doesn't do
It doesn't define the offer, doesn't know the customer better than the company and doesn't replace budget decisions. Nor should it publish anything without review.
Keep going
Questions
Frequently asked questions
What does an AI marketing agency do?
It applies artificial intelligence to parts of marketing that already exist: campaign analysis, ad variations, lead qualification, content drafts and follow-up automation. The goal is to gain time and consistency, with people reviewing what goes live and deciding the strategy.
How much does AI consulting for marketing cost?
In Oct/2026, Menthoria published an initial diagnosis at R$ 1,500 to R$ 5,000, a pilot project at R$ 3,000 to R$ 15,000 and follow-up at R$ 2,000 to R$ 8,000 per month. Corporate AI consulting, with complex data and integrations, is a different tier. These are Brazilian market references, not ArcWords prices.
What's the difference between hiring the service and taking an AI course?
A course teaches how to use tools. The service implements AI in your process, with integration, tests and an agreed success metric. Many companies take the course and never apply it because there's no time to design the flow. The two can coexist.
Does AI replace the marketing team?
No. It takes repetitive work out of the way. Strategy, offer, creative approval and budget decisions stay with people. The time freed up goes to what requires judgment.
Are licenses and APIs included?
Usually not. Tool subscriptions, AI API usage and cloud are paid by the company, directly to the providers. The proposal shows these estimated costs separately.
Does it work for small businesses?
Yes, as long as it starts with a concrete problem, such as answering leads faster or reducing reporting time. Small, measured projects work better than a full transformation all at once.
Next step
Let's talk about AI in your marketing.
Tell us where time is lost today: customer service, reports or content. The proposal comes after the diagnosis.